

In [4]:

    
from collections import Counter



In [5]:

    
Counter('with a string')









    Out[5]:





Counter({' ': 2,
         'a': 1,
         'g': 1,
         'h': 1,
         'i': 2,
         'n': 1,
         'r': 1,
         's': 1,
         't': 2,
         'w': 1})



In [6]:

    
Counter('with a string'.split())









    Out[6]:





Counter({'a': 1, 'string': 1, 'with': 1})



In [28]:

    
c = Counter([1, 1, 1, 2, 2, 3, 3, 3, 3, 3, 3, 3, 3, 5, 6, 100, 'test'])



In [9]:

    
c









    Out[9]:





Counter({1: 3, 2: 2, 3: 8, 5: 1, 6: 1, 100: 1, 'test': 1})



In [12]:

    
c.viewitems()









    Out[12]:





dict_items([(1, 3), (2, 2), (3, 8), (100, 1), (5, 1), (6, 1), ('test', 1)])



In [19]:

    
for k, v in c.iteritems(): 
    print "key:", k, "value:", v









    



key: 1 value: 3
key: 2 value: 2
key: 3 value: 8
key: 100 value: 1
key: 5 value: 1
key: 6 value: 1
key: test value: 1



In [21]:

    
c.most_common() # descending order of most common









    Out[21]:





[(3, 8), (1, 3), (2, 2), (100, 1), (5, 1), (6, 1), ('test', 1)]



In [22]:

    
c.most_common(1)









    Out[22]:





[(3, 8)]



In [23]:

    
c.most_common(3)









    Out[23]:





[(3, 8), (1, 3), (2, 2)]



In [34]:

    
c









    Out[34]:





Counter({1: 3, 2: 2, 3: 8, 5: 1, 6: 1, 100: 1, 'test': 1})



In [30]:

    
list(c)









    Out[30]:





[1, 2, 3, 100, 5, 6, 'test']



In [31]:

    
set(c)









    Out[31]:





{1, 2, 3, 5, 6, 100, 'test'}



In [32]:

    
dict(c)









    Out[32]:





{1: 3, 2: 2, 3: 8, 5: 1, 6: 1, 100: 1, 'test': 1}



In [39]:

    
c.most_common()[:-4-1:-1]









    Out[39]:





[('test', 1), (6, 1), (5, 1), (100, 1)]

defaultdict

The whole point is that it will always return a value even if you query for a key that doesnt exist. That value you set ahead of time is called a factory object. That key also gets turned into a new key/value pair with the factory object



In [41]:

    
from collections import defaultdict



In [42]:

    
d = {'k1':1}



In [43]:

    
d["k1"]









    Out[43]:





1



In [44]:

    
d["k2"] # this will get an error because the k2 key doesnt exist









    



---------------------------------------------------------------------------
KeyError                                  Traceback (most recent call last)
<ipython-input-44-d5a5b2fb6795> in <module>()
----> 1 d["k2"]

KeyError: 'k2'



In [45]:

    
d = defaultdict(object)



In [46]:

    
d['one'] # this doesn't exist, but in calling for it it will create a new element {'one' : object}









    Out[46]:





<object at 0x1023693f0>



In [48]:

    
d['two'] # same, this will add {'two' : object}









    Out[48]:





<object at 0x102369410>



In [50]:

    
for k, v in d.items():
    print "key:", k, "item:", v









    



key: two item: <object object at 0x102369410>
key: one item: <object object at 0x1023693f0>



In [59]:

    
e = defaultdict(lambda: 0) # lambda just returns 0 here



In [61]:

    
e['four']
e['twelve']









    Out[61]:





0



In [64]:

    
def error():
    return 'error'

f = defaultdict(error) #returned item must be callable or None



In [65]:

    
f['new']









    Out[65]:





'error'



In [66]:

    
f.items()









    Out[66]:





[('new', 'error')]

orderedDict

dictionary subclass that remembers the order items were added



In [70]:

    
d_norm = {}
d_norm['a'] = 1
d_norm['b'] = 2
d_norm['c'] = 3
d_norm['d'] = 4
d_norm['e'] = 5

# order isn't preserved since a dict is just a mapping
for k,v in d_norm.items():
    print k,v



In [73]:

    
from collections import OrderedDict



In [77]:

    
d_ordered = OrderedDict()
d_ordered['a'] = 1
d_ordered['b'] = 2
d_ordered['c'] = 3
d_ordered['d'] = 4
d_ordered['e'] = 5



In [78]:

    
for k,v in d_ordered.items():
    print k, v



In [79]:

    
from collections import namedtuple



In [81]:

    
# this is kind of like creating a new class on the fly
# the first parameter of a namedtuple is the name of the class/tuple type
# the second parameter is a space-delimeted list of properties of the tuple
Dog = namedtuple('Dog','age breed name')
sam = Dog(age=2, breed='Lab', name='Sammy')



In [82]:

    
print sam.age
print sam.breed
print sam.name









    



2
Lab
Sammy



In [102]:

    
Catzz = namedtuple('Cat', 'fur claws name')
mittens = Catzz(fur='fuzzy', claws='sharp', name='Mittens')
print type(mittens)
print type(Catzz)
print mittens[0]
print mittens.claws
print mittens.name
print mittens.count('fuzzy')
print mittens.index('sharp')









    



<class '__main__.Cat'>
<type 'type'>
fuzzy
sharp
Mittens
1
1



In [ ]: